Thorium : Is It an Alternative to the Current Uranium Nuclear Fuel Cycle?
- azizadussault
- 11 minutes ago
- 22 min read
Important note:
This synthesis does not cover all aspects of the Thorium cycle, notably everything related to radiotoxicity during the reprocessing of spent fuel to extract Uranium 233 (noted U233) and the manipulations within a complete cycle for the fabrication of a Thorium 232-Uranium 233 (noted Th 232-U233) fuel, which are also very serious (not to say redhibitory) additional obstacles for the implementation of a Th232-U233 cycle. Also not mentioned are all the technical problems specific to the types of nuclear reactor technologies described in this note.
In this synthesis, we place ourselves solely within a neutronic concept approach. This note is based on several public documents from scientific literature.[1]
The basic elements [2]
In nature, there exists only one fissile nucleus that can be used to produce energy in a nuclear reactor: it is the nucleus of the isotope Uranium-235 (the uranium atom comprises two isotopes in its natural state: U238 and U235 therefore two different nuclei—note that U238 is very weakly fissile). U235 is found only in a quantity of about 0.7% in uranium ore. Depending on the nuclear reactor types, U235 is therefore used either in its "natural" composition (types such as UNGG—Natural Uranium Graphite Gas—for example , French technology developed in the past but abandoned), or in an enriched composition—typically on the order of 4%—(types such as Pressurized Water Reactors PWR for example).
U235 has a high probability of fissioning with a "slow" neutron. But the neutrons resulting from this fission are fast, so they must be "slowed down": these are therefore reactor types using a moderator or "slower" ( such as water, heavy water, or graphite).
Uranium possesses, moreover, another very precious particularity, which is the presence of another isotope, U238 the most abundant at 99.3% in the natural state, this nucleus has the remarkable property of being a fertile nucleus. That is to say in a nuclear reactor, this fertile nucleus, by capturing a neutron emitted during the fission of the other nucleus U235, creates a new fissile nucleus: Plutonium-239[1]. We have thus created, in situ in the reactor, an artificial fissile nucleus which in turn will be able to participate in maintaining the chain reaction and therefore in the production of energy.
To fix ideas, when a nuclear fuel assembly is definitively discharged (one says "spent") from the reactor for disposal toward La Hague reprocessing plant, the energy it has produced during the some 3 to 4 years of its stay in the reactor comes for about 53% from the fissions of U235, 40% from the fissions of Pu 239 and the balance of 7% from the fissions of U238 (which also has the property of being "marginally" fissile for high-energy neutrons). And what is more, this in situ production of plutonium happens in a relatively short time through various decays, 2/3 days, so it is almost continuous.
This specificity of natural Uranium to possess "within itself" both a fissile isotope and a fertile isotope that transforms rapidly into a fissile isotope is at the base of the Uranium-Plutonium cycle. Moreover, when an assembly is completely "spent" in France (and it is the only country to do on an industrial scale), it is "reprocessed" in the La Hague plant in order to extract separately:
• on one hand the Plutonium (with an isotopic composition of isotopes Pu 239/240/241/242) and the residual uranium (called URT: uranium from reprocessing, containing still about 0.9% U235, which allows it to be re-enriched again into URE: re-enriched uranium). Plutonium is also used as a new fissile fuel to manufacture a MOX fuel (a mixture of depleted U with Pu) for the PWRs. Eventually, Pu can also be used to develop the line of Sodium Fast Reactor, particularly well-suited to "better use" and recycle Plutonium (this a long term issue).
• on the other hand, a mixture of waste (FP: Fission Products, and minor actinides: Neptunium, Americium, Curium) which are vitrified and constitute High-Level Long-Lived waste (HA-VL).
Note: Plutonium is fissile with "slow" neutrons (especially the odd isotopes), but also with fast neutrons (all isotopes combined, which is the basis of fast neutron reactor type, thus without a moderator). For reasons related to the neutronic properties of Plutonium, it is possible to use a MOX fuel in PWRs only once (mono-recycling of MOX). Indeed, spent MOX after a stay in a PWR contains too many even isotopes that are poorly fissile with slow neutrons. For this reason, spent MOX assemblies are currently temporary stored in pools awaiting future reprocessing and use in Sodium Fast Reactor where all isotopes of Plutonium have good fission properties.
But unfortunately, Thorium, present in the natural state, is composed almost exclusively of the isotope Th232 whose nucleus is not fissile, but which is a fertile nucleus, which by neutron capture and successive decays results in the production of another "artificial" fissile nucleus, U233 not present in nature (see the appendix which describes the Th2332-U233 chain). One will note the presence in this chain of an intermediate nucleus, Protactinium-233—particularly "troublesome" because it is a "big eater" of neutrons, which are thus lost for "fertilizing" the Th232 and again unfortunately it has a decay of 27 days, which maintains the production of U233 for a long time even if the reactor is stopped! This poses criticality problems that must be countered.
Consequently, to be able to benefit from this property of "fertility" of Th232 and establish a Th232-U233 cycle, it is first necessary:
to have an intermediate step to load this natural isotope Th232 into a "classic" nuclear reactor, for example, a reactor loaded with U235 (or Plutonium if available), to "have excess neutrons" in addition to the neutrons indispensable to maintain the chain reaction, in order to transform the Th232 into U233. "Excess" means that if, for example, 2.4 neutrons are produced by the fission of a fissile nucleus, it is necessary to "keep" one to fission the fissile nucleus (to maintain the chain reaction), one can lose 0.4 here, and there, but one absolutely needs an "excess" neutron to "fertilize" the fertile nucleus Th232 which, after some "adventures," will give a fissile nucleus U233.
then to treat the fuel in a specific plant to extract the U233 (an operation of great technological and radiotoxicity/radioprotection difficulties).
then eventually to manufacture a mix fuel Th232-U233 and inject it either into a classic reactor to minimize the necessary fissile load, or build a new reactor to "prime" the Th232-U233 cycle. This is only of interest if this new reactor is at a minimum iso-generator, or a breeder (iso- or breeding is the property of a reactor, when one discharges the fuel, to "recover" respectively as much or more fissile material—all fissile isotopes combined—than that put in at the start; otherwise one "dries up" the cycle, because one must always "inject" fissile material into the cycle).
One sees here an "original weakness" of Thorium, which is not possessing in its natural state, like Uranium, both a fissile isotope and a fertile isotope, but only a fertile isotope, which does not allow for a Th-U reactor type similar to the U-Pu type. This "original weakness" therefore obliges one to pass through a "complementary cycle" to Uranium or Plutonium to "prime" Th232-U233 cycle.
What is called a Th232-U233 cycle?
It is necessary to specify clearly what we are looking for. We disregard the time necessary for a complete fuel cycle, that is to say:
the cooling time of spent fuels after having produced energy in the reactor, before reprocessing,
the reprocessing time,
the fuel fabrication time,
and the operating time of the reactor.
All these durations total up to decades. We therefore place ourselves solely on the neutronic concepts. This being well-specified, one can define two aspects of the Th232-U233 cycle:
Either introduce Th232 (fertile isotope) into a reactor (classic U235 or Pu)—or into a fleet of reactors—and then extract the U233 from it, to reintroduce it into this classic reactor. This type of cycle has the objective of reducing, for a given energy production, the quantity U235 or Plutonium for the operation of such a fleet of reactors. This Th232-U233 cycle would therefore have interest in decreasing the necessary fissile loads (currently only based on U235 and Pu) to produce energy. Let us always keep in mind that in this path of the Th232-U23 cycle, a sufficient fissile load (called a nourishing core) is always needed (for example with U235 enriched up to about 20%) to have excess neutrons in addition to those that maintain the chain reaction to "fertilize" the Th232 and obtain U233. In this approach, we will evaluate which type of reactor allows the best conversion factor[1] thus, in the end, hoping to minimize the total quantity of fissile material to produce the same energy.
Or accumulate the U233 produced after reprocessing in a fleet of reactors (with nourishing cores) also loaded with Th232, to obtain a sufficient load of U233 to build a reactor loaded solely with Th232-U233 This option only makes sense if such a reactor has a conversion factor (capacity to produce more U233 than the initial load) large enough to obtain over time an equal load to build a new reactor and initiate the construction of a fleet over long periods. If the reactor is a sub-generator, then one "dries up" the quantities of U233 (one must always inject U233 to maintain this cycle), and if it is an iso-generator, then one recovers only one load to build a single additional reactor.
2.1. What are the reactor concepts loaded with Th232 in a cycle using "nourishing" core U235 or Pu to produce U233?
It is therefore necessary to verify technically the feasibility of U233 production and the conversion factor.
The path of Molten Salt Reactors (MSR) in slow neutrons with graphite as a moderator. The first idea in the 1950s was to use Th232 in MSR mixed in liquid phase with U235 fuel (salts based on Fluorine, or Lithium-7, or Beryllium, which are suitable because they are "low absorbers" of neutrons; thus the objective is to maximize the number of "excess" neutrons to "fertilize" the Th232) In fact such MSRs with graphite as moderator, was rather in an intermediate speed spectrum—called epithermal, i.e., "a bit faster"—having a better "eta" factor—see appendix—for a better conversion factor,.
The molten salt mixed with the fuel ensures both the role of fuel and coolant[1] Thanks to the liquid phase circulating in the reactor, it is possible to continuously extract a certain quantity to "send" it to a processing unit to "recover" the U233 and "rid" off the molten salt-fuel of "troublesome" isotopes (on one hand the fission products—notably Xenon, a "big eater" of neutrons—and on the other hand Protactinium Pa 233, a "neutronic poison," the "father" of U233 in the chain decay, but with a "long" period of 27 days, which leaves "time" in the reactor for a competition between the disappearance of Pa233 by "sterile" neutron absorption and its "useful" decay toward U233—see appendix).
One sees here the complexity of this concept which was experimented with by the Oak Ridge Laboratory in the USA from 1965 to 1969 (MSRE[2] Molten Salt Reactor Experiment of 8 MWth). Following the MSRE, an industrial breeder project, the MSBR (Molten Salt Breeder Reactor) of 1 GWe, was studied by the Oak Ridge Laboratory with an objective of a breeding factor of 1.07 (ensuring a doubling time of 10 years in theory to recover one equivalent U233 quantity than fissile Uranium burnt in the core, but given the losses, 20 years was announced), but at the cost of an imperative to process the entire molten salt-fuel in less than 10 days, which would have required a huge large-capacity processing unit. The project was stopped in 1976.
The path of Molten Salt Reactors in "fast neutrons": The idea is to remove, in the concept of slow neutron MSR, the graphite which is a source of difficulties, both on the level of the material itself which undergoes degradations requiring its replacement every 5 years approximately (long-lived radioactive waste whose disposal path remains a difficulty), and the proven risk of criticality runaway because of a positive temperature coefficient. The removal of graphite therefore leads to the Fast MSR concept which would use Pu as nourishing core and thus improve iso-generation (or even breeding). The inventory of fissile material in fast MSRs is very unfavourable. No experimental reactor has been built. The concept of a molten salt fast reactor was brought back to light and was integrated in the 2000s as one of the 6 concepts studied within the framework of the GIF (Generation IV International Forum). In France, it is the CNRS that is pursuing academic studies of this concept (with or without Thorium).
The path of Sodium-cooled Fast Reactors
This path would use a nourishing core made of MOX (Pu+depleted uranium) with Th232 blanket in order to fertilize it to U233. The main issue is that such blanket is solid thus having a permanent production of Protactinium Pa 233 which cannot be continuously evacuated. This leads to a very large nourishing core for having enough neutrons in excess for fertilizing the blanket. In addition there us competition in fertilisation of both U238 in the core and Th232 un the blankets.
India is working on this path and so far have built a Prototype Fast Breed reactor (PFBR of 200 Mwe) having reached recently criticality. But so far the blanket are with depleted uranium, objective being to operate the PFBR to get experience of Sodium Fast Reactor, then to move later to a Th232 blanket. As said, the necessary oversized nourishing core to produce power with Th232 blanket is very non-economically competitive. But for India, such PFBR is part of an overall long term strategy for using Thorium domestic mines, enabling to ban uranium import (see here below).
The path of PWR reactors. In the 1960s–1980s, the Th232-U233 cycle was experimented with in several light water reactors loaded with oxide fuel: Th-Uranium Oxide (based on U235 as a nourishing core), essentially to validate the concept of an operation where U233 was created in situ and to calculate the conversion factor. This path was explored in the Shippingport PWR demonstrator reactor in the USA from 1977 to 1982. It was a 60 MWe PWR using a mixed Th-U Oxide fuel. At the cost of great technological feats, necessitating the constant "rearranging" of the nourishing assemblies needed in the core to "counteract" the "parasitic" absorptions of neutrons and maintain the chain reaction and have excess neutrons, the Americans obtained a breeding factor of 1.014. With such a low breeding factor, the doubling time (time after which one "recovers" enough fissile material to start another reactor) is on the order of 50 years. This path was abandoned by the Americans because it was not viable on an industrial and economic level.
The path of Heavy Water Reactor: This path, thanks to heavy water being less of a "neutron absorber," was explored "on paper" by the Canadians in the years 1970–1976. They abandoned this path because the breeding factor, and thus the doubling time, was judged unattractive.
The path of High-Temperature Reactors (HTR): It has also been experimented with using highly enriched Uranium and specific carbide fuels. It is a reactor type with graphite as a moderator and CO2 (or helium) as a coolant. The use of graphite allows for the reduction of the enrichment of the U235 nourishing core. Achievements in HTR have been numerous:
Experimental Reactors: HTR AVR reactor in Germany (Jülich) of 15 MWe, which operated between 1967 and 1988, DRAGON HTR reactor in Great Britain (Winfrith) of 20 MWth, which operated between 1964 and 1975, Peach Bottom HTR reactor in the USA of 40 MWe, which operated between 1966 and 1974. These reactors used a core with a ThC2-UC2 fuel.
Then Prototype Reactors: THTR reactor in Germany of 300 MWe which operated between 1983 and 1999, Fort Saint Vrain HTR reactor in the USA of 330 MWe which operated from 1974 to 1989.
Despite the initial successes of the experimental HTR reactors, the prototype HTR reactors all experienced technical difficulties which have led to this day to the abandonment of this HTR line with this Thorium-Uranium cycle.
We have seen in this panorama that these different reactor technologies are more or less well-suited to producing U233 which would allow in theory for the constitution of a stock of U233 (by running reactors during decades) and the envisaging of a Th232-U233 reactor line ( as seen here above the doubling time is in the range of several decades).
Assuming having accumulated U233, what are the reactor technologies to ensure energy production and sustainability of U233 breeding ?
In this paragraph, we are therefore going to explore what types of reactors could operate with a fuel solely of Th232-U233 fuel cycle. It is therefore assumed, in theory, that sufficient quantities of U233 are available to envisage a new fuel cycle.
For such reactors, it is mandatory to design a breeder reactor, otherwise, the concept of this cycle loses its interest. For intrinsic neutronic reasons, this Th232-U233 cycle is better suited in a "slow" neutron reactor than in a fast neutron reactor.
By using Heavy water as moderator (thus in theory minimising fissile fuel), the Indians are developing an AHWR (Advanced Heavy Water Reactor) concept of 300 Mwe using the Thorium-232–Uranium-233 cycle (but it seems that this concept still requires an input of MOX-Th-Pu nourishing fuels ensuring reactivity to have a sufficient source of "excess" neutrons for the production of U233) There is to date neither a prototype nor an experimental reactor of the AHWR type.
This concept of AHWR is the last step for the Indians, of a very long-term strategy in 3 steps:
The first step is the production of Pu in heavy water reactors called PHWR (Pressurised Heavy Water Reactor[3]) in order to produce Plutonium (U-Pu cycle—the Pu forms in the Uranium-238 of natural uranium). This line does not require enriched uranium. This step then requires a fuel reprocessing plant of the La Hague type consisting of reprocessing the spent fuel from the PHWR reactors to extract the Pu and load a fast neutron reactor of step 2. India operates a fleet of 16 PHWR reactors of 220 MWe, 2 reactors of 540 MWe, and a unit of 700 MWe. Other 700 MWe PHWR reactors are under construction. This reactor type is of great complexity because it is necessary, continuously while the reactor is in operation, to replace the natural uranium fuel with a machine allowing for the maintenance of the heavy water under pressure in the tubes containing the fuel. Based on existing operation of PHWR’s, India has “extracted” sufficient quantity of Pu to initiate the step 2.
The second step is based on a Sodium-cooled Fast Neutron Reactors (similar to SuperPhenix in France) using a MOX fuel with Pu and U depleted. This second step is developed by the Bhabha Atomic Research Center, with the PFBR (Prototype Fast Breeder Reactor of 200 MWe). Recently, April 2026, PFBR loaded with MOX (U-Pu) and a blanket (at this stage of Uranium-238) has reached criticality. The objective is first to master the operation of the PFBR, confirm neutronic characteristics and calculation of breeding aspect with U238 at this stage? Then to envisage the use of Th 232 in the blankets and produce U233 In this case, this type of PFBR will probably require a larger nourishing core[4] (see here above as well).
And finally the third step is after having reprocessed the fuel from step 2 to extract the U233, the AHWR concept (Advanced Heavy Water Reactor) as described above, which could ensure their energy independence with their own Thorium, without resorting to the import of enriched Uranium-235 (having Plutonium produced in step 2, this can also be used in the nourishing core because it is needed). This is a very long-term plan of great complexity on the technological level and requiring many years of R&D, without any guarantee on the sufficient breeding capacity in this third step, which is the fundamental criterion to ensure the sustainability of the Th232-U233 fuel cycle.
This very long-term strategy has experienced many difficulties in the development of the PHWR line and the PFBR prototype, and requires an industrial tool of great complexity, and one can reasonably wonder if the Indians will be able to carry it through to the end.
Conclusion
At the end of this panorama, we have placed ourselves in a "neutronic" approach of what a Th232-U233 fuel cycle could be.
First by examining how to introduce thorium into the current uranium-235 cycle and use the produced U233 to minimize the global input of fissile material into the cycle?
Then by assuming the U233 is available, examining which reactor technologies using the Th232-U233 cycle could be implemented with an objective of sustainability of the fissile material? (However, an input of nourishing core U235 or Pu is probably always needed to ensure the operation of the reactor.)
To the first question, we answer that only the line of molten salt reactors could be used to prime a Th232-U233 cycle with an objective of minimizing the quantity of fissile material. This line is the subject of studies within the GIF of 4th generation reactors (since the 2000s), and at this stage, they appear to fall under academic studies without perspective of any experimental realization.
Let us emphasize however that in mid-2025, China (The Shanghai Institute of Applied Physics—SINAP—belonging to the Academy of Sciences in China) commissioned a 2 MWth molten salt reactor, the MSR-LF1[1] (Molten Salt experimental Reactor—Liquid Fuel), similar to the MSRE experimental reactor of Oak Ridge in the USA from the years 1965. The USA recently "declassified" the scientific information from the MSRE.
To the second question, we answer that we do not reasonably see an industrial and economic path for massive production of U233 an indispensable condition to imagine developing an industrial cycle based solely on Thorium-232–Uranium-233. In the past, several countries were able to produce in experimental installations the quantities of U233 necessary for the pursuit of the experimental and prototype reactor realizations (all nevertheless needed a nourishing core, the objective being to validate the concepts and the conversion hypotheses).
The general conclusion of the High Commissioner for Atomic Energy dating from November 2005 keeps all its relevance: "In the current French context, there is no strong motivation emerging for the implementation of a Thorium line. Although such a line possesses several interesting qualities in principle, none of them seems able to justify the technological and industrial break that would constitute the replacement of the Uranium fuel cycle by a Thorium fuel cycle."
Appendix
The radioactive chains of production of the 2 artificial fissile isotopes Pu239 and U233 from respectively U238 and Th232
These 2 chains are similar but include a difference of fundamental importance which harms the Thorium cycle.
One notes in the Uranium cycle that Pu239 appears" relatively quickly (successions of two decays with periods of 23 min then 2.3 days). Because of this, in situ, within the reactor a "source," via these short decays, "rapidly" feeds the production of "artificial" fissile nuclei of Pu239 with a low probability of seeing the precursor Np239 disappears via parasitic neutron captures. This in situ "source" is a very precious property in PWRs, because it allows an input of "artificial" fissile nuclei which participate in the energy production of the reactor. When a PWR assembly is discharged, after having stayed about 3 years in the reactor, the nuclei of Pu239 have contributed for about 40% to the energy produced by this assembly, the nuclei of U235 for 53%, and the nuclei of U238 for 7%.
On the other hand, in the Thorium cycle, U233 appears more "belatedly" (successions of 2 decays with a period of 22 min, then especially of 27 days). This difference is of importance. Indeed, the precursor nucleus of U233 the Protactinium Pa233, staying relatively long within the reactor (period of 27 days), its probability of disappearing by "parasitic" capture of a neutron is high, thus "drying up" the production quantity of U233 coming from the decay of Pa 233 In addition, by this parasitic capture, an "excess" neutron is lost for fertilizing the Th232 This intrinsic "weakness" of the Thorium cycle has 2 consequences:
on one hand not being able to obtain "easily" a breeding factor for the thorium cycle reactor (in certain reactor concepts, one must always bring in fissile material via a "nourishing core" for the chain reaction
on the other hand during the shutdown of the reactor for discharge, then the Pa233 only decays (there are no longer neutrons for its disappearance via "parasitic" captures) to produce U233 by decay, obliging the provision of additional absorbing rods to counteract the excess reactivity of this fissile nucleus.
2. Conversion factor and breeding
Refer to the definition of the conversion factor given in note 4.
Let be very careful on the “scientific definition” to avoid any wrong interpretation.
Taking a reactor under operating at the equilibrium state. This means that the cycle N+1 will be the same of the cycle N. Let calculate (for example on a complete fuel cycle: a reactor at full power during one complete cycle:
X is the total number of all fissile nuclei “destroyed” (either because having fissioned thus producing energy, or having capture a neutron for another phenomena different from fission it is a “sterile capture” ). In this category of fissile nuclei in a PWR for example, nuclei are U235 (“fissionable” and having “sterile capture” as well), Pu239 (at the equilibrium they are numerous, “fissionable and “sterile capture” as well), Pu 241 (at the equilibrium less than Pu239,there are “fissionable and “sterile capture”). Note that even Pu240 and Pu242 exist in a PWR, they are negligible to be accounted as “fissile nuclei” for PWR technology (with “slow neutron”).
x is the total number of nuclei interacting with neutron for “fertile capture” (thus leading to a fissile nuclei). In this category of nuclei in a PWR there are U238 (“fertile capture”) and Pu240 (existing but not fissionable in PWR but “sterile capture” for producing Pu241 which is fissionable in PWR).
For Sodium Fast Reactor, fissile nuclei “destroyed” are all plutonium isotopes.
Conversion factor is x/X.
This conversion factor is calculated by complex neutronics codes.
Of course depending on the technology of reactor: PWR, Sodium Fast Reactor-SFR- (with or without blankets with U238), Molten Salt Reactor (without or with Th232), etc breeding factors are very different:
· For PWR (900 MW) conversion factor is 0.61
· For SFR (Super Phenix 1200 MW with U238 blanket): 1.24
· For MSBR Project: 1.07
· For PWR with Th232/U233 cycle: 1.014
It is only SFR (with U238 blanket) which could have a high conversion factor. (significantly greater than 1: it is a breeding factor). But conversion factor may influence other data (such void coefficient). Thus targeting a conversion factor should be carefully assessed.
What is the use of the conversion factor.? For example, 1.014 means that, one will be able to have “fertile nuclei “representing 1.014 times “fissile nuclei destroyed”.
Thus assuming to reprocess spent fuel we will recover a certain quantity of new fissile nuclei (from processing fuel and eventually blankets with fertile material).
The doubling time T is then defined as the time at the end of which one "recovers" double the fissile material than that introduced into the cycle. On a theoretical level, this then allows the use of the fissile material thus accumulated to start a second reactor, while having enough fissile material to continue the operation of the first, and so on.
In our example: (1.014) T =2 which is a doubling time of about T =50 years (without counting the losses during the succession of operating cycles).
Similarly with 1.07 as breeding factor doubling time would be 10 years.
Of course this is theoretical as it is necessary to have a complete industrial reprocessing chain of spent fuel to extract U233 and minimizes loss of material in the reprocessing. USA at the time of MSBR they announced a realistic 20 years as a double time.
The calculation of the conversion factor must be assessed globally on a reactor concept (nuclear fuel of the core and eventually the "blankets" around the core made of "fertile" nuclei like U238 or Th232), but an industrial approach needs to take the real fuel management (length of cycle, how many unload fuel and reload fuel at each cycle, time for reprocessing, time to manufacture fuel, etc…). These are “complicated scenario calculations”.
What are the properties of fissile nuclei favouring breeding ?
Two aspects to consider :
1.During the fission of a fissile nucleus, one notes υ (Greek letter "nu") the number of neutrons emitted. This number depends on the fissile nucleus considered, and on the speed (thus its kinetic energy 1/2mv2) of the neutron "fissioning" the fissile nucleus. To simplify, one generally defines two domains for the energy of the "fissioning" neutron:
• the "thermal" domain (or slow neutrons) where one reasons for reactor physics calculations for an energy from 0 to 0.025 eV.
• the "fast" domain (or fast neutrons) where one reasons for reactor physics calculations between 0.025 eV and 10 MeV.
Values of υ for the Uranium-Plutonium and Thorium-Uranium cycles
Cycle | Uranium-Plutonium | Thorium-Uranium | ||
| 235U | 239Pu | 232Th | 233U |
υ « thermal » | 2,42 | 2,87 | 0 | 2,43 |
υ « fast» | 2,43 | 2,94 | 0 | 2,53 |
One notes that uranium, whatever its nucleus U235 or U233 has similar values of υ, on the other hand, the fission of Pu239 produces more neutrons; this is one of the preponderant factors for breeding.
2.There is competition, for a given fissile nucleus U235/PU239/U233 between the "probability" to "capture sterilely" a neutron (it is "a lost neutron" without a fission) and the "probability" to "fission by 'eating' a neutron" (certainly the 'eaten' neutron disappears but we have new neutrons emitted). These properties are measured with an equivalent of "probability" noted as "cross section" (σ Greek letter "sigma"), either:
• σc cross section of "sterile" capture
• σf cross section of fission.
The sum σa =σc + σf is the absorption cross section, sometimes noted as "disappearance"; it is indeed the "probability" for a neutron to disappear either by capture or by fissioning.
The unit of cross sections is the "barn" it is a unit of area [1]which is worth:
10-24cm2
It is not a probability but an equivalent. The ratio σf/σa measures the "probability" for a neutron absorbed in the fissile nucleus to cause a fission. This is what is useful. The quantity η= * σf/σa (Greek letter "eta") then measures the number of neutrons emitted per neutron absorbed (thus disappearing). It is this quantity that is going to be representative of the breeding capacity. All these quantities depend on the nucleus considered and on the energy of the neutron.
It is necessary that η be significantly greater than 2 to "hope" that a reactor is a breeder: indeed, one neutron to fission a fissile nucleus and maintain the chain reaction + one neutron captured by a fertile nucleus to produce a fissile nucleus and fractions of neutrons captured "sterilely" in the moderator (eventually), structural materials, and leaks out of the reactor.
[1] The cross section measures in a way a "probability" of interaction between a neutron and the geometric surface of its target, in this case a nucleus.
Values of η for the Uranium-Plutonium and Thorium-Uranium cycles
Cycle | Uranium-Plutonium | Thorium-Uranium | ||
| 235U | 239Pu | 232Th | 233U |
η « thermal» | 2,06 | 2,11 | 0 | 2,29 |
η «fast» | 1,89 | 2,33 | 0 | 2,26 |
One sees through this table for the Uranium-Plutonium cycle that in the fast domain, Pu239 has a value of η=2.33 which allows breeding. This is the specificity of fast neutron reactors.
In the thermal domain, U235 has an η=2.06 "just enough" to ensure the chain reaction in a PWR, to compensate for sterile captures and leaks, and produce fissile material thanks to the fertility of U 238 to produce Pu239 but it is impossible to be a breeder. This is the specificity of PWRs. R&D was implemented to design a PWR HFC (High flux of conversion) by diminishing water volume proportion vs fuel volume for “hardening flux” to have better U238 fertilization. But such R&D was not successful.
On the other hand, for the Thorium-Uranium cycle, either in the thermal or fast domain, U233 has an η allowing "in theory" breeding, better in the thermal domain than in the fast domain.
However, to go further in the analysis, it is then necessary to calculate the breeding on the reactor as a whole (including in dynamics especially for the thorium cycle because of the appearance of Protactinium-233, "neutron eater" and with a period of 27 days). One then notes for a slow neutron reactor a very tight balance regarding the breeding capacity due to the intrinsic weaknesses of the Thorium-Uranium cycle: presence of Pa233 "neutron eater," no fission of Th232 thus no additional neutrons, and more or less significant "sterile" captures depending on the moderator. In practice, the experimental reactors having "tried" the Thorium-Uranium cycle reached a low breeding factor: 1.014 for the Shippingport reactor in the USA (a 60 MWe PWR) which in theory leads to a doubling time of about 50 years.
[1] The note BN 3 563 "Thorium fuel cycle" by Dominique Grenèche has been used for this analysis.
[2] Neutronics is interested in the interaction within the nuclei of atoms. When one writes U235 for example, one speaks of the nucleus U235 (composed of 92 protons and 143 neutrons)
[3] The production chain of Pu239 since the capture of the neutron by U238 includes 2 intermediate "short" decays of 23 min and 2.3 days which in a way maintains "continuously" a production of a new fissile isotope. (see appendix).
[4] The conversion factor is a fundamental parameter that characterizes the capacity of a nuclear reactor to produce fissile material at the same time as it consumes it to produce energy. More precisely, it is the ratio between the total fissile nuclei produced in the reactor through “fertilization” from “fertile nuclei” during a given period (for example during an irradiation cycle at equilibrium) and the total fissile nuclei consumed in the reactor during this same period (for fission and “sterile capture” on such fissile nuclei). It is a complex calculation. If the conversion factor us greater than 1 such reactor is named ‘breeder”.
[5] In this concept, molten salt with liquid fuel is also the coolant. Thus always leads to a circuit external to the core with a thermal exchange with a secondary circuit. There is therefore always a significant quantity of molten salt and fuel outside the core, which increases the quantity of fissile material per unit of energy volume produced.
[6] The composition of Molten Salt with liquid fuel was: 65% Li7F, 29.1%BeF4, 5% ZrF4, 0.9 % UF4
[7] The term "Pressurised" indicates that there are horizontal tubes containing pressurized heavy water—the coolant—and fuel. These tubes are inserted into a horizontal network of horizontal tubes ("calandria") filled with heavy water as a moderator at atmospheric pressure.
[8] This PFBR is poorly suited to the production of U233 with Thorium blankets because there is competition between the neutron absorptions of U238 (in the MOX) and Th232 Thus, a large nourishing core is needed.
[9] The MSR-LF1 uses a Fluor-Beryllium molten salt (fertile blanket of lithium-beryllium fluoride with more than 99.95% Li7 with a liquid fuel based on zirconium and uranium fluorides (HALEU: 10% U235 and 50 Kg of thorium -It seems that HALEU was not as usual at 19.75% ) The moderator is graphite. It seems that there is an inline reprocessing unit to extract fission products, but that it is planned after 5-8 years of operation to stop the reactor and proceed to a complete reprocessing of the liquid fuel. Only molten salt and molten fuel reactors, using an inline and continuous treatment to "get rid" of Pa 233 as quickly as possible, could aim for breeding in the thermal, or even fast, domain
[10] The cross section measures in a way a "probability" of interaction between a neutron and the geometric surface of its target, in this case a nucleus.
Comments